Adult brain cancers (mostly gliomas) are the most severe forms of human malignancies, usually with a 50% fatality rate within the first year of diagnosis. Apart from potential involvement of occupational, infectious and environmental agents in the etiology of brain cancer, evidence for genetic associations with malignant glioma, especially glioblastoma multiforme (GBM), has also been documented in population-based studies. The objective of this epidemiological research initiative is to confirm observations in a cohort from the San Francisco Bay area that certain class I human leukocyte antigen (HLA) variants (alleles and haplotypes) can differentially correlate with the onset and progression of GBM. The goal of the proposed work is to examine HLA and related (neighboring) polymorphisms in a new cohort of 145 GBM patients and 145 control subjects, with 1:1 matching by ethnicity (European American), gender, age group and county of residence. Allelic variants of three genes (HLA-A, HLA-B and HLA-C) in the HLA class I region, one (HLA-DRB1) in the HLA class II region and another (MICA) located between class I and class II loci will be resolved by molecular techniques to facilitate epidemiological analyses in this study population. The primary aim is to determine if the previously identified HLA class I markers can be confirmed or refined by data from the new cohort. As a secondary aim, identical analyses will be extended to 100 patients with anaplastic astrocytoma (the second most common form of adult brain cancer) and 100 properly matched control subjects (all are European Americans) in order to evaluate the generalizability of findings based on GBM. Overall, by defining the role of highly polymorphic HLA class I molecules that are essential to cytotoxic T-lymphocyte function and natural killer cell activity, this study will help clarify the importance of immune surveillance and inflammation in the onset and prognosis of adult brain cancer in the United States. This population-based study will examine the consistency of genetic associations with glioblastoma multiforme, which is the most common and most severe form of adult brain cancer. In particular, through epidemiological analyses of new as well as existing data from carefully selected populations, this work aims to confirm and refine relationships of human leukocyte antigen (HLA) genotypes to adult brain cancer and to pave the way for further collaborative research. [unreadable] [unreadable] [unreadable]